Annual Review of Ecology and Systematics 5: 419-463 (1974)
Demography of Plants pp. 427-429
J. L. Harper and J. White

Juvenile Period


The ideal speedy life cycle might involve a seed that germinates to expose a green flower that immediately proceeds to leave several seeds that germinate without delay. The flower would need to photosynthesize sufficiently to stock the new seeds with reserves and to support a root for the necessary mineral uptake. This ideal does not exist and what we see in the array of higher plant forms is a variety of compromises in which precocity of reproduction is sacrificed to the growth of a vegetative structure. The shortest time from seed to germination to first seed produced is found in annuals of deserts (333), dunes (215) and cultivated habits.

Figure 1 Generalized scheme for the development of a plant population, showing important demographic events. Germination of-seeds is here equated with birth.

A number of annuals start flowering and set their first seed when very young and small; flowers are borne in the axils of early leaves. [Given the correct day, length, and temperature treatments, germinating seeds of the annual Chenopodium rubrum were induced to flower when the cotyledons had expanded and only two rudimentary leaves had formed (62).] New flowers continue to be formed as the plant grows and seed set (and dispersal) continues at an increasing rate through the growing season. Some species (e.g. many Veronica species; Poa annua, Senecio vulgaris) are well fitted to an unpredictable environment in which, if the season turns out to be short, some seed will have been produced; however, if the season turns out to be long it can be fully exploited. Such annuals have indeterminate growth systems and are not normally clocked into flowering by periodic stimuli; they are commonly weeds in the uncertain or complex rotations of horticulture. There is often no post-reproductive life and perhaps no intrinsic death process-life is cut short by the intervention of drought, the first killing frosts, or, as the environment deteriorates, the plant slowly rots away while still in the process of producing more flower buds (203).

In other annuals a relatively long period of vegetative growth precedes a rather sudden transition to the flowering and fruiting stage; this is optimal behavior in an environment of high predictability and characterizes the weeds of many arable crops in which the timing of seed release is just before harvesting, e.g. Agrostemma githago, A venafatua, Papaver rhoeas, and P. dubium. The change from the vegetative to the reproductive phase in these species is often rigidly seasonal and timed by photoperiod (267, 276). The growth forms are commonly determinate and death follows seed set as if the act of seed production was itself lethal. Annuals with longer life cycles, e.g. winter annuals, also biennials, have a more or less long vegetative stage that changes under periodic stimuli to a flowering phase; vegetative apices are converted to a flowering condition and a "big bang" of reproduction is followed by death.

It is questionable whether the length of the juvenile period has any significant effect on the potential rate of population growth in annual species. If germination and the growth of the next generation followed immediately on seed set the length of the juvenile period would be of profound importance. In practice most annual species are rigidly seasonal in their germination. Even among annual weeds that may set seed and germinate at almost any season in the year (e.g. Poa annua, Senecio vulgaris) there is some uncertainty whether the populations are composed of a variety of ecotypes with different seasonal behavior or of a single type that can repeat several generations in a year. This problem warrants much further study.

In biennials at least one year of vegetative growth is made before seed is produced. Evidence from density experiments with Digitalis purpurea (Oxley, in preparation) show that, under stress, flowering may not occur for several years; it would appear that a minimal rosette size must be achieved before flowering. In deliberately sown populations of Dipsacus fullonum (Werner, personal communication) no rosettes

flowered in the first year, a few flowered in the second year, and the majority in the third summer; by the fourth summer a few stragglers flowered and a few very large four year old rosettes remained which showed no signs of flowering at all. The life table data varied from field to field and there was no correlation between flowering and rosette size except that rosettes <15 cm diameter did not flower.

Theoretically, a strict biennial would be required to produce the square of the number of seeds of a comparable annual to achieve a comparable population growth rate, assuming no mortality. If the mortality risk is evenly distributed throughout life the same rule holds, but if there is a particularly high risk in the juvenile stages the annual experiences it twice in two years and the biennial only once. The biennial habit is therefore clearly favored when seedling establishment is risky, and the best return on seeds that germinate is obtained by living long and producing many seeds at the end of that life (116). If environmental hazards affect all stages of the life cycle equally, there would appear to be no special advantage in the biennial habit.


The juvenile period in herbaceous perennials appears to be very variable: e.g. Prunella vulgaris (2 yr); Potentilla erecta and Cirsium palustre (4 yr); Ranunculus auricomus, Chrysanthemum leucanthemum, Polygonum viviparum (5 yr); Ranunculus acer (6 yr); Trollius europaeus, Alchemilla vulgaris, Geum rivale (8 yr). These values were obtained from field records in Finland (174), but it is probably very dangerous to generalize these values from habitat to habitat for the same species. Very early juvenile phases of perennial grasses may live many months or even years (244) without appreciable increase in size. This is called "resistance to inanition" (44, 45, 54). The same is true of dicotyledonous seedlings of woodland that may survive several months of darkness and then resume vigorous growth when suitable light conditions return (135).

The juvenile period of geophytes varies from one to at least seven years (even under ideal horticultural conditions), depending on the minimal size necessary for flowering and on the rate at which this minimal size is achieved (86). The juvenile period may last only 1 yr in Freesia, Dahlia, Tritonia, Brodiaea, and potato, 1-2 yr in Gladiolus and Allium, 2-3 yr in Lilium, 3-4 yr in Crocus and Iris, 3-5 yr in hyacinth, 4-6 yr in daffodils, 4-7 yr in tulips. In many species the growth of clones may start before flowering, but this does not seem to be correlated with the length of the juvenile period. In tulips the long juvenile period is accounted for by (a) large minimal bulb size for flowering, (b) a comparatively small annual increase in bulb weight (100-200%), (c) a season of bulb growth of only about 6 weeks, and (d) a low proportion (60%) of total plant weight in photosynthetic tissue (86). In general, species with tubers or corms reproduce more precociously than those with bulbs and storage rhizomes. The life cycles and demography of natural populations of bulbous species might be a particularly rewarding study.

Information on the length of the juvenile period and the longevity of a variety of plant species is brought together in Figure 2. Clearly herbaceous perennials show very wide variation, but the data for perennial herbs are even less reliable than the comparable data for trees and shrubs. Several of the observations quoted are essentially anecdotal and lack the precision in recording which is given for Ananas, Anthoxanthum, Carica, and Guadua. Many of the records are from quite specific environmental conditions in localized sites. In some cases the life span is for genets, but in most there is no distinction made between genets and ramets. In clone forming species genet life span is potentially very long (see later under longevity).

As a group, orchids are notable for their very gradual vegetative development from seed, as they require a mycorhizal association for several months or years before autotrophic tissues are formed (some species remain permanently heterotrophic or saprophytic). Orchis ustulata has remarkably delayed development of leaves, none being formed until it has developed a rhizome 10-15 yr old. Consequently, many orchids do not flower for many years after germination, e.g. Listera ovata, Spiranthes spiralis, and Cypripedium calceolus which require 13-16 yr growth before flowering (300). Listera ovata has a life span of at least 40 years (307).

Figure 2 The relationship between prereproductive life (or juvenile period) and total life span for perennial plants. The data give the known normal life spans and not exceptional examples of great longevity. Each of four groups—perennial herbs, shrubs, coniferous trees, angiospermous trees—is separately numbered, counting cumulatively from left to right and from bottom to top of the graph; a few numbers are inserted as a guide. The names of the species are given in the Appendix, with sources indicated. The heavy lines roughly delineate each group of species.


The length of the juvenile period for a number of woody species is given in Figure 2. These represent the period of juvenile, vegetative growth for "normal" trees, probably biased towards minimal age estimates, although a number are minimal ages for commercial seed production. There is a very general relationship between life span and the age of first reproduction, but the variation is large. Shrubs have a reasonably defined and circumscribed variation, and trees fall into two fairly distinct groups, angiosperms and conifers, the latter having earlier reproductive ages for a given length of life. Several conifers that may survive over 200 yr reproduce before age 10. Most angiosperms with a life span of over 200 yr spend at least 20 yr in the juvenile phase. The length of the juvenile phase and the longevity of the species are much more closely related in angiosperms than in conifers (the ratio is ~10:1 between these parameters) (note particularly the data for Quercus and Carya which illustrate this generalization). Very few hardwood trees live more than 400 yr and there are no records of trees in normal circumstances taking much more than forty years to reach reproductive age. These observations add greater precision to the general conclusions of Molisch (193): (a) plants with brief youth (period from germination to first seed production) have short longevity; (b) plants with a long youthful period generally enjoy great longevity; (c) a long youthful period is usually followed by a long and often very extended bearing time.

Of 57 species, hybrids, and varieties of Pinus observed in California (253) 55 produced ovulate cones at an average minimum age of 5.2 yr and 39 produced staminate cones at an average minimum age of 4.4 yr. Many of these data were from single individuals of a species and under the conditions of a forest genetics institute, not a commercial forest. This clearly shows that the common belief that timber tree species require 20 or more years to reach reproductive maturity does not apply to this genus. A rare example of a population study on the distribution of juvenile periods was made on Pinus sylvestris in Pennsylvania (97): of 1050 trees grown, 2 carried ovulate cones at age 3, 3 at age 4, 41 at age 5, 46 at age 6, 377 at age 7, and 790 at age 8. Only 199 had staminate cones by age 8, so that apparently the juvenile period for cones is shorter than for pollen. Very few data exist for tropical species (61, 197); the first recorded flowering of 50 dipterocarpous species in Malaya ranged from 17 to 36 yr and for most between 20 and 30 yr.

During the juvenile stages of development a tree commonly experiences ecological conditions markedly different from those in the mature phase. Perhaps associated with this is a frequently different morphology during the juvenile period (2, 274). Some parts of the subpopulation of a single tree may retain the juvenile characteristics while others develop into an adult and reproductive condition ["Topophysis" (193)], e.g. the adult and juvenile stages of the shrub Hedera helix (72). The juvenile period in Hedera lasts at least 10 yr (90) and (like plagiotropic shoots of Araucaria) the adult characters are retained permanently after vegetative propagation. The physiologic changes that occur before an adult condition is reached in woody plants are still not understood, and phrases like "maturation" or "phase change" (158) are only blankets for ignorance. The essential demographic feature of the juvenile phase is the absence of sexual reproduction. A certain minimum size may be necessary before flowering can begin (158). The first trees to flower at a given age are usually among the largest and most vigorous (275). In general, those environmental conditions (weather, soil, cultural practices) that promote vegetative growth tend to shorten the juvenile phase (130, 318). The length of the juvenile phase in apple and pear seedlings is negatively correlated with trunk diameter, and apparently genetic control of the growth rate determines when the minimum size for reproduction is reached. The length of the juvenile period is also correlated with parental characters such as the season of flowering and fruit ripening and the time from flowering to fruit maturity (319-322).

The genetic control of the length of the juvenile period has been clearly established. The period between planting and flowering in the oil palm (Elaeis guineensis) has been reduced by breeding from 45 to about 30 months (25). Three generations of selection for early flowering in Betula verrucosa reduced the juvenile period to two years (293) and precocious flowering in this species is apparently determined by a single gene (141). Precocity has obvious economic value in fruit tree culture and also in practical forestry because it increases the number of generations available for selection (263, 343).

The juvenile period can sometimes be modified by special physiologic treatment. For example, that of Betula verrucosa was reduced from 5-10 yr to little over a year in some plants by growing seedlings continuously under 18 hr photoperiods at 60-70°F (175); also in a remarkable experiment Pharis & Morf (218) reduced the juvenile period of Sequoia and Sequoiodendron seedlings from 20-70 yr to 8 and 12 months with gibberellin sprays. The very precocious Cupressus arizonica with a juvenile period of 18 months could be made more precocious (3 months) by gibberellin sprays (219). Surprisingly, juvenile periods cannot be reduced by grafting young tissue on to older stocks (72, 258).

In comparisons between species vegetative vigor is associated with late flowering. Within the species the more vegetatively vigorous individuals flower first. A generalization that describes much of the variation between species in the length of a juvenile phase is that shade intolerant, colonizing (r-type) species tend to have precocious reproduction associated with large seed numbers, small seeds, and high reproductive efficiency. In contrast, species that occupy later positions in forest succession and are shade tolerant have fewer and larger seeds and a long juvenile period. All available assimilates in early life are channelled into establishing canopy height and achieving competitive dominance. The fact that the length of the juvenile period is heritable presumably implies that the length to which the species has become adjusted in nature represents an adaptive response.

Continue reading